Zijun (Zizi) Gao

My research develops trustworthy and efficient statistical methods for causal inference in modern data settings, with a particular focus on heterogeneous treatment effects. When treatment effects vary across individuals, natural questions arise on adaptive trials, benefiting subgroups, and personalized treatments. These tasks further raise selective inference and multiple testing challenges due to the adaptivity of experiments and the multiplicity of subgroups.

I am also genuinely interested in the interplay between machine learning and causal inference: modern machine learning methods, such as neural networks and ML-assisted optimal transport, can help address causal problems with complex data; in turn, causal inference can contribute to model evaluation and to explainability of black-box algorithms.